---

name: github-automation description: "Routes OpenHands GitHub Actions examples, prompt-driven agent runners, TODO scanning, and example-run reporting workflows." disable-model-invocation: true metadata: disco-role: operating license: MIT

GitHub Automation

Use this sub-skill for SDK-powered GitHub Actions examples and supporting scripts: prompt-driven task runners, TODO scanning, example-run summaries, and workflow-level automation.

What this route owns

• Prompt-loaded agent runner scripts for local or CI use.
• TODO scanner behavior and filter rules.
• Example-run report rendering.
• GitHub Actions examples that invoke the OpenHands SDK.

Start here

Read references/actions-workflows.md for the workflow patterns and script entry points. Read references/troubleshooting.md when prompts, credentials, remote URLs, or example summaries fail.

Scripts:

• scripts/run_prompt_task.py
• scripts/todo_scanner.py
• scripts/render_examples_report.py

Typical triggers

• "Run an agent from a GitHub Action prompt file or URL."
• "Scan the repo for TODO(openhands) comments."
• "Render a markdown summary of example script results."
• "How do I wire this into a workflow dispatch job?"

Cross-links

• For local agent lifecycle and Conversation behavior, go to ../agent-core/SKILL.md.
• For remote server startup used by examples, go to ../remote-runtime/SKILL.md.
• For repository-maintenance policies or test selection, go to ../repo-development/SKILL.md.